Article Text

Original research
COVID-19 international experience in paediatric patients with congenital heart disease
Free
  1. Mary J Yeh1,
  2. Lisa Bergersen1,
  3. Kimberlee Gauvreau1,
  4. Oliver M Barry2,
  5. Sarosh P Batlivala3,
  6. Elsa Bjornlund1,
  7. Brian Boe4,
  8. Luiz Fernando Caneo5,
  9. Ulisses A Croti6,
  10. Thomas Doyle7,
  11. Shumaila Furnaz8,
  12. Rodrigo Cesar Moraes9,
  13. Michael L O'Byrne10,11,
  14. Erica de Oliveira Paes12,
  15. Alexis Palacios-Macedo13,
  16. Dimitar Pechilkov14,
  17. Nestor Fernando Sandoval15,
  18. Supratim Sen16,
  19. Mila Stajevic17,
  20. Marco Antonio Ferreira Travessa18,
  21. Kathy J Jenkins1
  1. 1Department of Cardiology, Boston Children's Hospital, Boston, Massachusetts, USA
  2. 2The Congenital Heart Center, New York Presbyterian/Morgan Stanley Children’s Hospital, New York, New York, USA
  3. 3Department of Pediatrics, Cincinnati Children's Hospital Medical Center Heart Institute, Cincinnati, Ohio, USA
  4. 4The Heart Center, Nationwide Children's Hospital, Columbus, Ohio, USA
  5. 5Department of Cardiovascular Surgery, Instituto do Coracao do Hospital das Clinicas de Universidade de São Paulo, São Paulo, Brazil
  6. 6Department of Pediatrics and Pediatric Surgery, CardioPedBrasil Children’s Heart Center, Hospital da Criança e Maternidade FUNFARME/FAMERP São José do Rio Preto, São Paulo, Brazil
  7. 7Division of Pediatric Cardiology, Monroe Carell Junior Children's Hospital at Vanderbilt University Medical Center, Nashville, Tennessee, USA
  8. 8Research Department, National Institute of Cardiovascular Disease, Karachi, Pakistan
  9. 9Dr. Carlos Alberto Studart Gomes Hospital, Hospital de Messejana, Fortaleza, Brazil
  10. 10Division of Cardiology, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
  11. 11Department of Pediatrics, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
  12. 12UTI Cardiopediátrica and Neonatal, Hospital do Coração (HCor), São Paulo, Brazil
  13. 13Department of Pediatric Cardiac Surgery, American British Cowdray Medical Center IAP, Mexico City, Mexico
  14. 14Department of Pediatric Cardiology, Multiprofile Hospital for Active Treatment National Cardiology Hospital, Sofiâ, Bulgaria
  15. 15Department of Cardiac Surgery, Fundación Cardioinfantil de Bogotá, Bogotá, Cundinamarca, Colombia
  16. 16Department of Pediatric Cardiology, Narayana Health SRCC Children's Hospital, Mumbai, Maharashtra, India
  17. 17Department for Pediatric Cardiothoracic Surgery, Mother and Children Health Institute, Belgrade, Serbia
  18. 18Department of Cardiovascular Surgery, Hospital Gaspar Vianna, Belem, Brazil
  1. Correspondence to Dr Lisa Bergersen, Department of Cardiology, Boston Children's Hospital, Boston BCH 3215, Massachusetts, USA; lisa.bergersen{at}childrens.harvard.edu

Abstract

Objective As COVID-19 continues to affect the global population, it is crucial to study the impact of the disease in vulnerable populations. This study of a diverse, international cohort aims to provide timely, experiential data on the course of disease in paediatric patients with congenital heart disease (CHD).

Methods Data were collected by capitalising on two pre-existing CHD registries, the International Quality Improvement Collaborative for Congenital Heart Disease: Improving Care in Low- and Middle-Income Countries and the Congenital Cardiac Catheterization Project on Outcomes. 35 participating sites reported data for all patients under 18 years of age with diagnosed CHD and known COVID-19 illness during 2020 identified at their institution. Patients were classified as low, moderate or high risk for moderate or severe COVID-19 illness based on patient anatomy, physiology and genetic syndrome using current published guidelines. Association of risk factors with hospitalisation and intensive care unit (ICU) level care were assessed.

Results The study included 339 COVID-19 cases in paediatric patients with CHD from 35 sites worldwide. Of these cases, 84 patients (25%) required hospitalisation, and 40 (12%) required ICU care. Age <1 year, recent cardiac intervention, anatomical complexity, clinical cardiac status and overall risk were all significantly associated with need for hospitalisation and ICU admission. A multivariable model for ICU admission including clinical cardiac status and recent cardiac intervention produced a c-statistic of 0.86.

Conclusions These observational data suggest risk factors for hospitalisation related to COVID-19 in paediatric CHD include age, lower functional cardiac status and recent cardiac interventions. There is a need for further data to identify factors relevant to the care of patients with CHD who contract COVID-19 illness.

  • COVID-19
  • Heart Defects, Congenital
  • Risk Factors
  • Global Health

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author, LB, upon reasonable request.

This article is made freely available for personal use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.

https://bmj.com/coronavirus/usage

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • To the authors’ knowledge, there has not been a multicentre study focusing on paediatric patients with COVID-19 and congenital heart disease (CHD) in an international setting, reporting results for low-income, middle-income and high-income environments.

WHAT THIS STUDY ADDS

  • This study reported data from a large, international cohort, including low-income, middle-income and high-income countries, of paediatric patients with CHD with COVID-19 which demonstrated an association between the need for hospitalisation related to COVID-19 infection and lower functional cardiac status, age and recent cardiac interventions.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • These findings allow healthcare providers to better counsel paediatric patients with CHD about the risks associated with COVID-19 infection for their population.

Introduction

The impact of the COVID-19 pandemic across the globe has highlighted the importance of rapidly collecting real-world data on the impact of COVID-19 in children with conditions that might result in higher risk of morbidity and mortality. While initial concerns centred around respiratory comorbidities, subsequent information has shown the significant impact of COVID-19 on all organ systems.1–6 Little data have been reported on COVID-19 in paediatric patients with congenital heart disease (CHD), particularly as paediatric patients were initially considered less likely to contract COVID-19 and at lower risk for severe clinical symptoms.

As of 2017, nearly 12 million people worldwide were estimated to be living with CHD,7 and many CHD manifestations leave little room for further deterioration due to COVID-19. Therefore, data on COVID-19 are crucial to informing care pathways for patients with CHD. Further, the pandemic has exposed healthcare disparities broadly, and significant disparities already exist for CHD in the USA and globally. By capitalising on networks from two pre-existing CHD registries, the International Quality Improvement Collaborative for Congenital Heart Disease: Improving Care in Low- and Middle-Income Countries (IQIC) and the Congenital Cardiac Catheterization Project on Outcomes (C3PO), this observational study aims to describe the risk factors of moderate and severe COVID-19 in paediatric patients with CHD in a diverse, international cohort.

Methods

Site recruitment

All participants of two collaboratives were invited to collect information on all patients with COVID-19 and CHD in 2020 at their respective institutions. IQIC, launched in 2008, includes 64 centres and collects data on CHD surgeries and catheterisations performed in low-income and middle-income countries (LMICs). C3PO, established in 2007, collects data on all CHD catheterisations performed at 17 US institutions. Data collection was permitted under existing IQIC agreements. Interested US C3PO institutions amended data use agreements to allow data collection about COVID-19, allowing for rapid deployment of a new data collection tool. A total of 29 IQIC centres and 6 C3PO centres elected to participate in this new initiative, collecting data on all patients with CHD with COVID-19 at their institutions, irrespective of previous cardiac intervention history. Prospective data collection began in May 2020, with all previous cases meeting inclusion criteria entered retrospectively. Patients were not involved in the study design or conduct.

Population

Centres reported data for all patients under 18 years of age with a previous CHD diagnosis and known COVID-19 illness during 2020 identified at their institution. All patients who were suspected to have COVID-19 due to clinical assessment or positive PCR test were included.

Web-based tool

The pre-existing IQIC (iqic.chboston.org) and C3PO (c3po-r3.chboston.org) web pages were modified to include a direct link to the password-protected data collection tool, which was built using the REDCap platform and launched on 8 May 2020.

Data capture

Sites confirming full data capture or verifying no COVID-19 and CHD cases assessed in 2020 were included. The REDCap platform required users to mark each section (medical history, testing information and clinical course) ‘Incomplete’ or ‘Complete’ for each patient to verify complete data capture. Cases marked ‘Complete’ for all sections were included in the analysis.

Patient characteristics

Patient characteristics included age, weight, sex, genetic syndrome (online supplemental table 1), presence of major active non-cardiac illness (online supplemental table 1), current oral medications (online supplemental table 1) and CHD diagnosis (online supplemental table 2). CHD status was described along two axes. First, CHD diagnoses were classified into the three categories according to anatomical complexity (simple, moderate or complex) adapted from adult classifications (online supplemental table 2).8 Second, clinical cardiac status was described using a novel variable currently used by the C3PO and IQIC-Cath databases composed of physiological and imaging criteria dividing subjects into three categories (good, moderate and poor) based on most recent echocardiogram information, expected haemodynamic data and baseline oxygen saturation (table 1). Due to differences in physiology, criteria differed for single-ventricular versus biventricular circulation (table 1). Patients with characteristics of multiple categories were assigned to the highest applicable classification of good, moderate or poor functional physiology status. Patient medical history included whether the patient had undergone cardiac catheterisation or surgery within the last 90 days, summarised as recent cardiac intervention, or if the patient had a future planned surgery or catheterisation (Y/N/to be determined).

Table 1

Clinical cardiac status components

To determine overall patient risk of significant COVID-19 illness (low, moderate or high), we considered definitions of anatomical complexity, clinical cardiac status and presence of a genetic abnormality using recommendations from the European Society of Cardiology (ESC) amended for paediatric CHD.8 Simple anatomical diagnoses and good cardiac status were low risk, moderate complexity and moderate cardiac status were moderate risk, and complex heart disease and poor cardiac status, high risk. If a genetic abnormality was present, the patient was classified as at least moderate risk, similar to ESC classification.8 Patients were categorised based on highest risk designation, so a patient with low anatomical complexity but poor cardiac status would be considered high risk.

COVID-19 data

Date of first positive PCR test, reason for PCR test (symptomatic, preprocedure screening, known/suspected contact, or other), or reason not tested were collected. COVID-19 symptoms were collected using a checklist based on the Center for Disease Control Symptom List as of May 2020.9 Other symptoms were collected as a free text field then grouped and summarised as congestion, malaise, fatigue, gastrointestinal problems or other. Date of symptom onset, date of symptom resolution if known, and date of most recent clinical information were recorded.

Patients were stratified by highest level of care received: non-hospitalised, hospitalised non-intensive care unit (ICU), or ICU level care. Last-known patient clinical status was captured as symptomatic, no symptoms or deceased. If a patient required hospitalisation for COVID-19 illness, additional data were collected including major clinical manifestations of COVID-19 (significant arrhythmia, decline in ventricular function, renal insufficiency, coagulopathy, or other major clinical manifestations) and use of COVID-19 specific treatments (hydroxychloroquine, immunoglobulins, immunomodulators (including steroids), antiviral treatment (eg, remdesivir), or other COVID-19 specific treatment). In addition, need for supplemental oxygen, need for mechanical ventilation, length of hospital admission and length of ICU admission were collected for hospitalised patients.

Analyses

Demographic and COVID-19 illness characteristics were summarised using frequencies and percentages for categorical variables and medians, IQRs and ranges for continuous variables. Summaries were performed for the entire cohort, stratified by non-US versus US centres and by patients who were not hospitalised versus hospitalised non-ICU versus admitted to the ICU. Logistic regression was used to investigate predictors of admission to the hospital and admission to the ICU for the following patient characteristics: age, non-cardiac comorbidity, genetic syndrome, recent four interventions, anatomical complexity, clinical cardiac status and risk category. ORs were estimated with 95% CIs. Variables significant at the 0·20 level by the likelihood ratio test were considered for inclusion in multivariable models; forward selection was used and p<0·05 was required for retention in the final models. Discrimination was assessed using the c statistic. Analyses were performed in Stata V.16 (StataCorp, College Station, Texas, USA).

Results

Cohort characteristics

A total of 35 sites reported 339 COVID-19 cases in patients with CHD<18 years of age in 2020. Twelve non-US and 6 US centres had at least one case, and 17 non-US centres confirmed no cases identified at their institution during this time period (online supplemental figure 1), (online supplemental table 3). There was no notable regional clustering of centres that did or did not report cases. Cohort characteristics are summarised in table 2.

Table 2

Status at the time of COVID-19 diagnosis and COVID-19 testing data

Anatomical complexity and clinical cardiac status

Using the highest applicable category of anatomical complexity, clinical cardiac status and genetic syndrome in 279 patients with all three components recorded, 143 (42%) patients were categorised as low risk, 98 (29%) as moderate risk and the remaining 98 (29%) patients were high risk (table 2, table 3). This classification method characterised the non-US population as majority high risk, with 60% of patients categorised as high risk compared with 19% in the US population. Functional univentricular patients accounted for 13% of the cohort, most of whom were older patients who had undergone a Fontan palliation. Specific cardiac diagnoses are detailed in figure 1.

Table 3

Overall risk score

Figure 1

Number of patients by cardiac diagnosis not requiring hospital care (grey), requiring non-ICU hospital care (blue) and intensive care (red). CHD, congenital heart disease; ICU, intensive care unit; PGE, prostaglandins; s/p, status post.

COVID-19 testing

Among the cohort, the indication for most patients who tested positive was COVID-19 symptoms (table 2, figure 2), with fever and cough as the most commonly reported symptoms. Only three non-US patients were tested exclusively because of known/suspected contact, while 44 US patients were tested only for known/suspected contact (table 2, figure 2). There were six patients whose COVID-19 was not confirmed by PCR, all from the USA, who were not tested due to lack of test availability at the beginning of the US pandemic.

Figure 2

Relative frequency of highest level of care provided or death by risk factor.

Outpatients

Of the 339 patients reported with COVID-19 illness, 255 (75%) were not admitted to the hospital admission for COVID-19 symptoms. Characteristics of hospitalised and non-hospitalised patients are shown in table 4. Fifty-one per cent of the patients who were not hospitalised tested positive for COVID-19 but had no reported symptoms of illness (table 4).

Table 4

Patient characteristics by highest level of care provided

Hospitalisations

Eighty-four patients received hospital care for COVID-19, detailed in table 4. Of the 84 hospitalised, 44 were admitted to non-ICU care while 40 received intensive care (12% of all reported cases). Characteristics of ICU and non-ICU patients are shown in table 4 and stratified by US and non-US in online supplemental tables 4 and 5, respectively.

Functional univentricular heart was reported in nine patients and was the most common CHD diagnosis, followed by Tetralogy of Fallot, pulmonary valve abnormalities and isolated septal defects for patients hospitalised without ICU level care (figure 1, online supplemental table 6).

Patients receiving general floor care were symptomatic for a median of 14 days and hospitalised for a median of 4 days, with eight (18%) hospitalised prior to symptom onset (table 4). Median length of hospitalisation for patients receiving ICU level care was 17 days, with a median ICU stay of 8 days (table 4). Shortness of breath was reported in 30 (70%) ICU patients, fever was reported in 29 (67%) patients, and there were no reports of chills or loss of smell or taste in ICU patients.

The most common major clinical manifestation in ICU patients was a decline in ventricular function (23%), with six (15%) patients experiencing a new significant arrhythmia (table 4). New supplemental oxygen requirements were reported in 34 (85%) ICU patients, with 21 patients requiring mechanical ventilation (table 4). Use of COVID-19 specific treatments was uncommon in US patients (online supplemental table 4). In the non-US population, 12 (48%) patients were placed on immunomodulators and 10 (40%) patients used hydroxychloroquine to treat the COVID-19 illness (online supplemental table 5). Overall, 20 patients who required ICU care had no symptoms at follow-up, with 13 (33%) patients reported as deceased (table 4).

Deaths

Eighteen deaths were reported (5%), with 15 (83%) in the non-US cohort, 4 of which were in asymptomatic neonates. The 14 patients with COVID-19 symptoms are detailed in table 5.

Table 5

Patient characteristics and COVID-19 manifestations and treatment in deceased patients with CHD diagnosed with COVID-19

Patient risk factors and outcomes

The highest level of care required for specific anatomical diagnoses is shown in figure 1. The highest level of care and mortality from COVID-19 by anatomical complexity, clinical cardiac status and overall risk are shown in figure 2.

In univariate analysis, age under 1 year, recent cardiac intervention, anatomical complexity—complex classification, clinical cardiac status and overall risk were significantly associated with need for hospitalisation and ICU admission (p<0.05) (online supplemental table 7, figure 2). Presence of a non-cardiac comorbidity and genetic syndrome were not statistically significantly associated with hospital or ICU admission (online supplemental table 7). Composite patient risk accounting for patient anatomy, physiology and genetic syndrome produced a value of p<0.05 for hospitalisation and ICU admission while clinical cardiac status produced a value of p<0.001 (online supplemental table 7, figure 2). Clinical cardiac status correlated most strongly with death (figure 2).

In a multivariable model, clinical cardiac status and recent cardiac intervention remained statistically significant predictors of both outcomes (table 6). The c-statistic was 0.75 for hospital admission and 0.86 for ICU admission (table 6).

Table 6

Multivariable model for admission to hospital and admission to ICU in paediatric patients with congenital heart disease and COVID-19

Discussion

This study leveraged two existing multicentre CHD registries to describe COVID-19 illnesses in 2020 from 35 institutions in 14 countries, providing an observational snapshot of COVID-19 in paediatric patients with CHD in diverse populations. Severe COVID-19 illness in paediatric patients with CHD was associated with poor baseline clinical cardiac status and recent cardiac intervention. Anatomical diagnosis did not have a significant relationship with hospitalisation or intensive care. Symptomatic patients reported a median symptom duration of 2 weeks. A quarter of reported COVID-19 diagnoses required hospital admission for COVID-19 symptoms, and patients requiring higher levels of care for COVID-19 illness often presented with fever and shortness of breath.

This study used a composite metric to assess overall cardiac status. Clinical cardiac status showed the strongest correlation with highest level of care needed, with 60% of patients with poor cardiac status requiring admission to the ICU and 83% of patients with good cardiac status not requiring any form of hospitalisation (figure 2, online supplemental table 7). Cardiac diagnosis or anatomical complexity did not correlate with need for hospitalisation or ICU admission (figure 2, online supplemental table 7). The data regarding baseline clinical status are not surprising as respiratory pathogens are more likely to affect patients with lower baseline health status. However, the data regarding anatomical complexity are interesting as increasingly complex CHD lesions are typically associated with suboptimal outcomes compared with simpler forms of CHD. This is presumably because SARS-CoV-2, a primarily respiratory pathogen, may affect the functional status and symptom severity of the patient independent of underlying anatomical CHD complexity.

Recent cardiac intervention was also associated with hospitalisation in our multivariate analysis. While children who recently underwent an intervention may have slightly lower baseline functional status as they had a haemodynamically significant CHD lesion that required an intervention, this finding may allow providers to better counsel patients undergoing interventions. In addition, recent cardiac intervention remained statistically significant in the multivariate model after accounting for the patient’s clinical cardiac status.

Previous studies of patients with structural, functional and rheumatic heart disease demonstrate the importance of resource availability and environment on patient outcomes.10 11 However, the implications of severe respiratory disease can differ greatly between patients with structural heart disease, patients with rheumatic disease and electrophysiology patients, which may dilute the ability to determine clinical risk factors. This study focuses on structural CHD to explore the specific risk factors that may be associated with severe COVID-19 disease.

US-based investigations reported increased hospital morbidities in COVID-19-positive children and adults with moderate or severe CHD,12 while another associated advanced physiological stage with moderate or severe infection.13 At this time, the authors are unaware of other studies comparing international outcomes of paediatric patients with CHD and COVID-19. When comparing the US and non-US populations, non-US patients that required hospitalisation were more likely to be younger and have more severe illness. Again, this is in keeping with indications for hospitalisation among respiratory illnesses. These patients were also primarily tested only when symptomatic or for preprocedure testing, while US patients were also tested for known contact with someone infected with COVID-19.

Recent advancements in care for complex CHD have increased the percentage of these patients surviving beyond infancy, with deaths decreasing by 35% from 1990 to 2017.7 However, this shift is driven by countries with advanced healthcare systems. Mortality and morbidity due to CHD remain highest in LMICs, where data are more difficult to acquire, leading to large uncertainty intervals. Furthermore, as countries shift towards improving health systems and economies, non-communicable diseases contribute more heavily towards infant mortality, increasing the visibility of CHD in such environments.

Our study shows that a large percentage of COVID-19 among hospitalised patients in LMICs was hospital acquired (33%) (online supplemental table 5). In addition, the burden of COVID-19 in LMICs remains high as a result of low vaccination rates.14 In the USA, vaccine use in paediatric patients has recently been authorised. Future work will be able to assess the impact of vaccination on outcomes and international availability and use of paediatric vaccinations in patients with CHD. These confounding factors, and the undue burden of both COVID-19 and CHD in LMICs, highlight the critical need to ensure equitable access to COVID-19 vaccinations, personal protective equipment, widespread availability of SARS-CoV-2 real-time PCR testing and general resources for all international populations, especially LMICs.

Limitations

Limitations of this study include burdens of data entry during a global pandemic, and differences in screening strategies among centres/countries, which has the potential to introduce site and patient selection bias. This study only includes cases identified by the participating centre as part of their routine clinical care. All patients who previously or currently receive care for CHD at the institutions were not contacted. As an observational study, follow-up data were limited to that regularly collected by the institution. The US emphasis on widespread testing of COVID-19 identified several outpatient cases with minimal follow-up data regarding symptoms and duration. To partially mitigate discrepancies in test utilisation among regions and during different eras of the pandemic, we collected reason for testing. Thus, while a higher percentage of cases in this study were reported in the USA, testing was performed for reasons other than patient symptoms. We also observed a difference in ICU patients with recent cardiac interventions by cohort. The reason for the differences in rates of recent cardiac intervention in ICU patients between LMICs and high-income countries (HICs) is not clear but may be confounded by differences in preprocedural testing and hospital-acquired infections. For these reasons, the primary data of this study are descriptive. Due to variation in resource availability, we chose to stratify the population by HIC and LMIC in supplemental tables but not overstate the comparative implications.

In addition, the level of care provided to patients who were hospitalised prior to COVID-19 diagnosis may demonstrate the level of care required unrelated to their infection and reflects the severity of other medical issues (eg, CHD) and not COVID-19 illness. We attempted to mitigate this by collecting information on hospitalisations necessitated by COVID-19 illness through careful wording of registry questions. However, we are limited in the conclusions regarding escalation of care or ultimate impact of care on outcomes for patients hospitalised prior to COVID-19 illness.

Conclusion

Data regarding clinical information and outcomes for patients with CHD with COVID-19 infection have been limited. Using international multicentre data from two CHD registries, our data demonstrated an association between the need for hospitalisation related to COVID-19 infection and lower functional cardiac status, age and recent cardiac interventions. For patients with CHD with COVID-19, the incidence of hospitalisation, severe COVID-19 disease burden, and mortality are concerning. Further data and multinational collaboration can help identify additional factors relevant to the care of paediatric patients with CHD who contract COVID-19.

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author, LB, upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

Institutional Review Board approval was obtained at Boston Children’s Hospital.

Acknowledgments

The authors thank Kaitlin Doherty-Schmeck and Lauren Gunnelson for their assistance in the design and coordination of this study.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Correction notice This article has been corrected since it was first published to correct affiliation 17.

  • Contributors MJY drafted the initial manuscript, assisted in study design and collected data. LB assisted in study design, developing data collection tools and provided guidance in drafting the manuscript. KG guided and conducted all statistical analyses. OMB, SPB, BB, LFC, UAC, TD, SF, RCM, MLO’B, EdOP, AP-M, DP, NFS, SS, MS and MAFT all assisted in local data collection and provided critical reviews of the manuscript in the drafting and revision stages. EB assisted in data collection and revision of the manuscript, and site coordination. KJJ guided study design, development of the data collection tool, and provided expert oversight in drafting, revising and finalising the study and manuscript, and serves as the guarantor for this work.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Map disclaimer The inclusion of any map (including the depiction of any boundaries therein), or of any geographical or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.